The role of shifts in delay of reinforcement, incentive quality and incentive quantity in the frustration effect

Abstract

The present study sought to investigate the effects of shifts in delay of reinforcement, incentive quality and incentive quantity on the Frustration Effect (FE). To this end, 164 naive female albino rats were given 42 training and 36 test trials in an L-shaped double runway under 22 1/2 hr. food and water deprivation and under a 24 hr. intertrial interval. During training three conditions existed in G1 and Ss received either 16% (Group 16), 2% (Group 2), or nonreward (Group 0). All Ss received 2% in G2 throughout the experiment. Prior to the test series, Ss in each of the sucrose groups (Group 16 and Group 2) were randomly assigned to six subgroups receiving the following conditions in G1: 2% with no delay, 2% with a 15 sec. delay, 16% with no delay, 16% with a 15 sec. delay, nonreward with no delay, and nonreward with a 15 sec. delay. Group 0, which was given nonreward in G1 during training, continued on the same condition during the test series. Starting and running time measures were recorded in each runway and were converted into speed scores (ft./sec.). Separate analyses were performed on the speed scores over Trials 31-42 of the training session, and over Trials 1-12 and Trials 25-36 of the test session. In the training phase, the starting and running speeds for Group 16 were significantly faster than Group 2 and Group O in the first alleyway (A1) and the starting speeds for Group 16 were significantly slower than Group 2 and Group O in the second alleyway (A2). In the test session, introduction of delay of reinforcement significantly decreased A1 starting and running speeds and increased A2 starting speeds over Test Trials 25-36. No significant interaction was obtained, however, between the delay variable and incentive variable when the latter was held constant during the training and test trials. Removal of the incentive quantity in G1 resulted in significantly slower A1 starting and running speeds over both Test Trials 1-12 and Test Trials 25-36 for the 16% incentive and over Test Trials 25-36 for the 2% incentive and faster A2 starting speeds over Trials 25-36. The performance, in the latter case, however was not reliably different from the performance of the control group (Group D). Incomplete reduction of incentive in G1 did not significantly affect either A1 or A2 performance. finally, an increase in incentive quality in G1 failed to affect A1 performance but did result in faster A2 starting speeds during the later stages of the test session. The theoretical implications of the results were discussed.